Institutional investors, corporate governance and firm performance in an emerging market: evidence from Vietnam

Abstract This study examines the relationship between institutional investors, corporate governance, and firm performance in Vietnam. The findings on Vietnamese listed companies indicate that while institutional investors are less likely to hold shares of companies with larger board sizes, Chief Executive Officer (CEO) duality, and ultimate control by the state (except for state-owned institutions’ perspective), the effect of their ownership on firm performance depends on whether they are pressure-sensitive (grey) or pressure-insensitive (independent) institutions. In Vietnam, independent institutional investors monitor the company and their investment more effectively than grey institutional investors. They can significantly influence management decisions and improve shareholder value. In contrast, grey institutional ownership is either negatively or insignificantly related to firm performance due to conflicts of interest, as they have a potentially related business relationship with the invested companies.


PUBLIC INTEREST STATEMENT
Corporate governance is always a topic that attracts a lot of researchers, especially in emerging markets like Vietnam where corporate governance reports have only been required to be published in recent years. This paper analyzes the relationship between institutional investors, corporate governance and performance of listed companies in Vietnam. In particular, the research analyzes the different influences of financial institution shareholder groups including grey, independent, private and state on factors in corporate governance, business characteristics and firm performance. Research results show that institutional shareholders prefer to invest in businesses with good corporate governance quality, large scale, and long history. Meanwhile, the influence of institutional ownership on firm performance depends on different categories of institutions.

Introduction
Corporate governance is no longer strange for businesses, investors, or stakeholders. An enterprise with an excellent corporate governance system is considered an attractive market magnet. Corporate governance includes principles and policies that help businesses orient, operate, and control; hence, it is considered a valuable mechanism for improving firm value (Gompers et al., 2003). Among the different factors, the ownership structure is a critical variable in corporate governance studies because it significantly influences important decisions in companies (Zattoni, 2011). The linkage between corporate governance, ownership structure, and firm performance has attracted much attention from researchers (Kumar & Zattoni, 2015;Lemmon & Lins, 2003;Utama et al., 2017;Zheka, 2005). Previous studies conclude that corporate governance quality and firm performance depend on different ownership structures. Shleifer and Vishny (1997) mentioned that the role of major shareholders is undeniable in corporations. Among them are institutional shareholders operating in financial sectors such as commercial banks, investment banks, securities companies, and insurance companies. Other investors can run their businesses in the fields of production or manufacturing.
Various studies have examined the determinants of firms' institutional ownership (Dahlquist & Robersson, 2001;Ferreira & Matos, 2008;Gompers & Metrick, 2001) and the effect of institutional investors on firm valuation and performance. Cornett et al. (2007) point out that institutional investors with no potential business relationship with the companies in which they invest (independent institutional investors) positively influence firm value. In contrast, institutions with some business relationships with firms invested (grey institutional investors) show no effect on firm performance. The findings of Ruiz-Mallorquí and Santana-Martín (2011) on Spanish firms and Muniandy et al. (2016) on Australian firms are also consistent with Cornett et al. (2007). Guo and Platikanov (2019) employed publicly traded Chinese firms from 1999 to 2010 to examine the determinants of various large institutional investors and listed companies' corporate governance characteristics. They find that institutional ownership is positively associated with firm value and that the effect is mainly driven by the ownership of independent institutions rather than privately owned institutions. The question of interest is whether different types of institutional investors influence corporate governance and firm performance in a less developed market like Vietnam, where issuing corporate governance reports has recently become mandatory for listed companies. The existing literature using the Vietnamese context has examined the effect of one or two types of institutional ownership of Vietnamese listed companies at a time (e.g., Ha & Hiep, 2019 on domestic institutional ownership and foreign institutional ownership), and there has been no research investigating the other types, such as state and private institutional investors or grey and independent investors, as well as their preferences on firm characteristics and their effects on firm value. Therefore, this study examines the relationship between institutional investors, corporate governance, and firm performance, mainly how institutional investors' preferences affect listed companies' business structures and characteristics in Vietnam. Following prior studies (Cornett et al., 2007;Guo & Platikanov, 2019), we classify institutional shareholders into two dimensions: (1) grey institutions versus independent institutions and (2) state-owned versus privately owned institutions. Our sample includes 409 listed companies in Vietnam from 2010 to 2019. The results show that while institutional investors are less likely to hold shares of companies with larger board sizes, the duality of the Chief Executive Officer (CEO), and ultimately control by the state (except for state-owned institutions), the effect of their ownership on firm performance depends on whether they are grey or independent institutions. In Vietnam, independent institutional investors monitor the company they invest in more effectively than grey institutional investors.
This study contributes to the literature on institutions' preferences for business structure and characteristics and how institutional ownership affects firm value in the context of Vietnam as an emerging market. First, it provides more evidence on the choice of institutional investors for listed firms based on corporate governance characteristics and business structure, especially in countries in emerging markets such as Vietnam. Research results show that institutional investors in Vietnam prefer to invest in listed companies with good corporate governance quality, such as smaller board size and separation of CEO and chairman roles, as well as in firms with large-scale, older age, and VN30 members. We also applied PCA analysis to check the robustness of the results. Second, some studies have examined the influence of ownership structure on firm performance in Vietnam. However, our results are slightly different from those of previous studies, indicating that the impact of institutional ownership on firm performance depends on the type of institution. This is the first study to classify institutional ownership in Vietnam into grey and independent institutions, in which independent institutions have positive effects on firm performance. Finally, this study has implications for managers of listed companies and regulators. Listed companies can determine suitable percentages and types of institutional ownership that enhance firm performance, and investors can have a broad knowledge of corporate governance to make decisions. Other stakeholders, such as stock market entities and the government, can offer more suggestions for improving the stock market.
The remainder of this paper is organized as follows. Section 2 presents previous studies and develops the hypotheses. Section 3 describes the data collection process, construction of variables, and empirical models. Section 4 discusses the empirical results and the endogeneity treatment. Section 5 presents conclusions and recommendations.

Institution's preferences on corporate governance features
First, we use Giannetti and Simonov's (2006) and Gompers and Metrick's (2001) theories to determine the different types of institutional ownership. Giannetti and Simonov (2006) conclude that major shareholders greatly influence voting in managerial decisions, but not all shareholders have an incentive to do this actively. Although security benefits accumulate for all shareholders, most shareholders involved in company management are interested in personal benefits. Therefore, the quality of corporate governance leads to different investor preferences. Giannetti and Simonov's (2006) research shows that domestic and foreign, institutional, and small individuals who enjoy only security benefits are reluctant to invest in companies with weak corporate governance in Sweden. By contrast, investors who may extract private benefits (i.e., large domestic individual investors) do not avoid companies with weak corporate governance. Li et al. (2006) investigated the relationship between institutional ownership and corporate governance in China. The findings of this study indicate a significant linkage between institutional ownership, CEO duality, and board composition. These features are classified as components of corporate governance in the companies. Accordingly, Duc and Thuy (2013) argue that board size, female board members' presence, CEO duality, board members' education level, board directors' working experience, independent (outside) directors' presence, board compensation, board ownership, and blockholders are generally the elements of corporate governance from a sample of 77 listed companies in Vietnam.
Additionally, the state retains its ultimate control over many listed companies in Vietnam. Guo and Platikanov (2019) argue that privately owned companies can achieve superior performance to state-owned companies because the privileges given to internal state shareholders drive managers to expropriate wealth from other stakeholders. Therefore, the discussion above motivates us to investigate the first hypothesis on institutional investors' preferences for corporate governance characteristics.
Hypothesis 1. Institutional investors in Vietnam are less likely to hold shares of firms with larger board sizes, CEO duality, and, ultimately, control by the state.

Institution's preferences on business structures and characteristics
Being prudent is essential for institutions worldwide, particularly in Vietnam, as Gompers and Metrick (2001) mention. They point out that institutional investors must be cautious when dealing with the legal environment. Therefore, firms with good liquidity, a long-established history, and a reputation in the market are more attractive to institutional investors. This study expects institutional ownership to be positively related to firm size, firm age, and members of the VN30 index. Consequently, except for corporate governance features, we develop a second hypothesis on institutions' preferences for firm characteristics.
Hypothesis 2. Institutional investors in Vietnam are more likely to invest in firms with a larger size, older age, and members of the VN30 index.

Grey institutions and independent institutions
In addition to investigating Vietnamese institutional investors' preferences for choosing businesses based on corporate governance and firm characteristics, we continue to analyze the effects on the firm performance of Vietnamese listed companies when there is investment from institutional investors. A company benefits from having major shareholders as its driving force to increase firm value (Shleifer & Vishny, 1986). However, having large shareholders incurs certain costs for corporations, stemming from their motivation to exploit the benefits of other stakeholders. Cornett et al. (2007) argue that pressure-sensitive (grey) institutional investors (e.g., banks and insurance companies), being shareholders, might have potential business relations with the firms, due to which they could be reluctant to create challenges to management decisions.
Grey institutional investors will likely affect management by sustaining and expanding their nonequity-related business relationships. The dominance of business relationships could be over their interests as shareholders and might not necessarily align with others' interests. For instance, banks are typically attentive to assuring extended debt obligations and thus are interested in encouraging their borrowers to undertake safer investment projects with consistent cash flows. Banks could deliberately steer management away from investment projects with a higher net present value (NPV) but a higher risk of protecting their loans. They prefer lower risk with lower NPV projects, resulting in a devaluation of firm value.
By contrast, pressure-insensitive (independent) investors are more incentivized to monitor and control their business activities. Following Guo and Platikanov (2019), we classify our sample into two groups: grey and independent institutional investors. Grey financial institutions include banks and insurance companies. Independent financial institutions consist of secure and venture capital firms. Therefore, we tested the third hypothesis as follows: Hypothesis 3. The effect of grey institutions on firm performance is negative, while that of independent institutions is positive.

Privately-owned institutions and state-owned institutions
After "Doi Moi" in 1986, Vietnam's economy underwent significant restructuring. However, state ownership still plays an essential role in the ownership structure of Vietnamese listed companies, and the state retains ultimate control over many partially privatized companies, especially in leading industries that significantly impact the country's economic and national security foundation. Before 2012, there were still listed companies in Vietnam, state ownership rate of which was over 90%. Given the importance of major institutional shareholders, it is necessary to investigate whether state-owned or privately owned institutional investors have different effects on firm performance. We expect both privately owned and state-owned institutions to affect firm performance positively; however, the effect of privately owned institutions could be greater than that of state-owned institutions. Therefore, we propose the fourth hypothesis as follows: Hypothesis 4. The positive effect of institutional ownership on firm performance is more significant for privately owned than state-owned institutional ownership.

Data
The initial sample for this research consists of all Vietnamese companies listed on the Hanoi Stock Exchange and Ho Chi Minh Stock Exchange from 2010 to 2019. These datasets are publicly disclosed in companies' annual and corporate governance reports following the regulations of the State Securities Commission. The data are obtained from Fiingroup and Vietstock, two reliable organizations providing data related to Vietnamese listed companies. Observations without sufficient information to test the hypotheses are excluded from the sample. In addition, all data were winsorized at the 1% level to control for outliers. As a result, 4090 observations were adopted as the final sample of this study, equivalent to 409 listed companies.

Institutional ownership measures
This study uses the information provided in the annual reports of Vietnamese listed companies to determine institutional ownership for the tests. Under Article 6, Sub article 9 of the Vietnam Securities Law (2006), the majority shareholder (or significant shareholder) means a shareholder directly or indirectly owning at least five percent or more of the voting stocks of an issuing organization. The percentage of institutional ownership of Vietnamese public firms in the research data is extracted from the list of large institutional shareholders, as only organizations and individuals that become the majority shareholders of a public company shall report their stock ownership to the authorities (Article 29, Sub article 1, Vietnam Securities Law 2006). Therefore, in this study, we only use large institutions as institutional investors extracted from the Vietstock organization's report. Institutional ownership represents the ratio of the number of shares held by institutions to the number of outstanding shares held by the company.
Following Guo and Platikanov (2019), we manually identify the financial institutions among large institutional shareholders and classify them into the following groups: • State-owned institutions (Starate) include financial institutions owned by the state or state legal entity.
• Privately owned institutions (Pirate) are identified as financial institutions not owned by the state or state legal entity.
• Grey or pressure-sensitive financial institutions (Greyrate) comprise banks and insurance companies.
• Independent or pressure-insensitive financial institutions (Indrate) consist of securities companies and venture capital firms.
Under Articles 114, Sub article 1.a, Vietnam Securities Law 2006, each ordinary share of a listed company should carry one vote. Therefore, it can be assumed that the ownership percentage closely reflects the voting rights held by investors in Vietnamese listed companies.

Corporate governance
Previous studies have examined the structure and efficiency of corporate governance systems.
Many studies have pointed out the critical role of the board of directors and have recognized this role as a mechanism that enhances corporate and economic performance. These arguments are addressed empirically using a sample of US firms and find that having a small board enhances a company's performance and positively influences investors' behavior and company value (Yermack, 1996). In addition, Horváth and Spirollari (2012) prove that board size influences CEO compensation incentives, as compensation programs represent an essential responsibility of the board of directors, and companies with oversized boards tend to become less effective. The Administrative Council indicates the board of directors in Vietnam. Therefore, this study includes the number of members of the Administrative Council (Adnum) to control institutional investors' preferences regarding the size of the board of directors.
CEO duality is also considered to be a measure of corporate governance quality. CEO duality refers to a situation in which one person can simultaneously hold the CEO and chairperson of the board of directors. Findings from previous studies of CEO duality are also controversial. Bhagat and Bolton (2008) proved that corporations with CEO duality could reduce business performance. On the other hand, some researchers have argued that the role separation between the CEO and chairman could split strategic decision-making and policy implementation, thereby increasing agency problems between senior management and directors. Therefore, to control institutional investors' preferences on the role separation of CEO and Chairman, this study includes a dummy variable (COB) equal to one if the same person holds the CEO and the Chairman positions and zero otherwise. Bai et al. (2004) find that the presence of large shareholders is positively associated with the firm's Tobin's Q. Institutional investors might gain better protection in companies where the top largest shareholders closely monitor each other. This study includes the number of large institutional shareholders (Insnum) as a measure of internal corporate governance to control company ownership structure.
The state still plays a vital role in listed companies in leading industries in Vietnam. Shleifer and Vishny (1997) argued that a state-owned company's managers might seek internal benefits that lead to inefficiencies in the firm's operations. Additionally, property rights theory argues that fully-privatized firms would perform better than government-controlled firms, as the power of control and decisions allows private shareholders to act towards maximizing shareholder wealth (Alchian & Demsetz, 1972). Accordingly, whether the state has ultimate control over a listed company might be a vital characteristic for performance-driven institutional investors. This study constructs an indicator variable (STATE), taking a value of one if the ultimate controlling shareholder is the state and zero otherwise.

Other control variables
The dependent and independent variables constructed in this study are summarized in Table 1. This table presents the expected outcome of each independent variable.
In Vietnam, the VN30 index tracks the total performance of the top 30 large-cap liquid stocks listed on the Ho Chi Minh Stock Exchange, along with two popular indices in Vietnam: the VN and VN30 indices. This index was developed by Phoenix Global Wealth Management-a division of the Phoenix Capital Group-in affiliation with the S&P Dow Jones Indices on 6 February 2012. This study includes a dummy variable (VN30) equal to one if a company is a member of the VN30 index in a specific year and zero otherwise to measure institutional investors' preference for index membership.

Tobit model
Institutional ownership, as the dependent variable, is censored. Petersen (2008) argued that this data type must be modified explicitly. Therefore, institutional investors' preferences for corporate governance features of listed companies in Vietnam are estimated by applying the Tobit model.
The dependent variables-institutional investors-include different categories of institutions (grey, independent, privately owned, and state-owned institutions), while the explanatory variables to measure a firm's corporate governance quality are Adnum, COB, and Insnum. Accordingly, firm size (SIZE), firm age (AGE), and firm's presence in the VN30 index (VN30) are included to identify institutional investors' prudence in holding stocks with lower risk and higher liquidity. Regarding other firm characteristics, equation (1) consists of TobinsQ, STATE, return on assets (ROA), firm leverage (LEV), annual sales growth (Sgrowth), and firm's market share (Mkshare) as control variables. All estimations consist of time-fixed effects (�φ t ). Consequently, Tobit estimations are as follows:

Pooled ordinary least square model
To test the effect of institutional ownership on firm performance, we apply the pooled ordinary least squares model with Tobin's Q as the dependent variable. We include the number of members on the board of directors (Adnum), a dummy variable for CEO-Chairman duality (COB), and the number of large institutional investors (Insnum) as internal corporate governance measures. According to Alghifari et al. (2013), this study consists of return on assets (ROA), as it significantly affects Tobin's Q. Moreover, following Guo and Platikanov (2019), specific firm characteristics are investigated, such as firm size (SIZE), leverage (LEV), and sales growth rate (Sgrowth), which are consistently related to firm performance. Market share (Mkshare) is also included as a control variable for performance (Graves & Waddock, 1994). Finally, the research adds up the control variables STATE since there is still an unusually high percentage of companies ultimately owned by the state in Vietnam.
In Model (2), we have two groups representing the institutional ownership percentage from two mutually exclusive groups of investors. Group1 indicates independent (Indrate) and grey (Greyrate) institutional investors, while Group2 indicates privately owned (Prirate) and state-owned (Starate) institutional investors. All estimations contain time-fixed effects (�δ t ). The pooled OLS estimations are as follows:

Three-stage least square model
Following previous studies, the relationship between institutional ownership and firm performance is likely to be mutually interdependent. Therefore, the three-stage least-squares (3SLS) method was used to eliminate potential endogeneity because it allows for cross-correlation between the equations.
Besides the simultaneity and endogeneity between institutional ownership and firm performance, these equations directly compare the marginal effects on firm performance for each mutually exclusive investor group. Sets of excluded variables were used to identify the parameters. Moreover, since each equation in the 3SLS model requires at least one specific exogenous variable, total institutional ownership is included only in Tobin's Q equation. Accordingly, each type of institutional ownership (Insnum1 and Insnum2) is included only in the institutional ownership equations. Tobin's Q regression is the primary test of interest in these systems of equations. The other two regressions on Group1 and Group2 were adopted to overcome the endogeneity issue.
Group1 and Group2 in models (3), (4), and (5) have the same classifications as in Model (2). The parameters were estimated using the three-stage least-squares method. All estimations contain time-fixed effects (�δ t and�θ t and�μ t ). The 3SLS estimations were constructed as follows:

Descriptive statistics
In Regarding the two variables measuring firms' profitability, Tobin's Q has a larger difference than ROA, with standard deviations of 0.493 and 0.096, respectively. While ROA has a minimum value of −1.587, showing that in the research samples, there are still inefficient businesses and negative after-tax profits, Tobin's Q has a range of variation from 0.145 to 9.044, showing that in terms of the market index, firms in the sample are all potentially profitable. This is because ROA uses book value, while Tobin's Q represents the expected value of the business. In addition, the variable Sgrowth with a minimum value of 0.003 also proves that the listed companies have a specific growth in annual sales. Firm size is measured by the SIZE variable, with a relatively large standard deviation of 0.657, indicating that firms have differences in market capitalization. This can be explained by the fact that businesses in many industries have different sizes. In addition, firms in the observed sample also have a market share compared to the industry from small to large, as shown by the variable Mkshare, with the smallest value of 0.012 and the largest value of 0.806. For the variables AGE and LEV, representing firm age and financial leverage, there is not much variation between the minimum value (0.301; 0.124) and the maximum value (1.875; 3.491), showing that firms in the sample do not have much difference in the number of years of establishment as well as the problem of financial risk. However, only 17.1% of the companies belonged to the VN30 group. This is understandable because VN30 includes firms with significant market capitalization on the HOSE in Vietnam.
Finally, corporate governance measures such as the number of board members (Adnum), CEO duality (COB), number of institutional shareholders (Insum), and state ownership (STATE) show that companies in the sample have board members ranging from 3 to 11, of which institutional shareholders have up to seven members. Up to 45.5% of the samples are state-owned companies due to the equitization program in Vietnam for more than ten years. However, the variable CEO duality (COB) has a standard deviation of 0.453, indicating that 45.3% of companies have a chairman on the board who is also a Chief Executive Officer (CEO). This is also a problem in the corporate governance system.

Institutional investors' preferences
First, it examines corporate governance features and firm characteristics that can attract institutional investors. The results from the Tobit model are reported in Table 3 with the dependent variables: grey, independent, privately owned, and state-owned institutional ownership percentages.
In Table 3, the results reveal that the coefficients of the COB are significantly negative for almost all types of institutional investors at high significance levels (1% and 5%, respectively), except for the insignificant coefficient of COB in Column (1) (Greyrate). This finding suggests that institutional investors in Vietnam prefer listed firms with separate roles as the CEO and chairman of the board of directors. Moreover, the coefficients of Adnum are insignificant in both Columns (1) and (3), but they are significantly negative in Columns (2) and (4) at the 5% and 10% significance levels (independent and state-owned institutional ownership). Additionally, the coefficients on Insnum indicate that institutional investors are likely to invest in companies with a robust internal monitor, while those of Adnum point out the preference on a small scale of the board of directors. This result supports Hypothesis 1, which states that institutional investors in Vietnam are less likely to hold firms with larger board sizes, CEO duality, and state controls ultimately. These characteristics also represent good quality corporate governance. Therefore, this finding aligns with that of Guo and Platikanov (2019). However, grey and state-owned institutions are still positively related to companies ultimately controlled by the state.
The results for Hypothesis 2 are also presented in Table 3, which indicates that institutional investors in Vietnam are more likely to invest in firms with larger sizes, older ages, and members of the VN30 index. The positive and statistically significant coefficients on SIZE and AGE in three out of four regression results are consistent with Hypothesis 2: institutional investors demand more mature firms to be prudent, except for those on AGE in Column (3) (privately owned institution). Being prudent is essential to institutional investors (Gompers & Metrick, 2001), hence they prefer to hold outstanding shares of safe companies. These results are consistent with Guo and Platikanov's (2019) findings. This argument suggests that privately owned institutions are more likely to participate in new investment opportunities than are state-owned institutions. The positive and significant coefficients of VN30 in column (3) are also associated with Hypothesis 2. The coefficients of SIZE, AGE, and VN30 in column (1) (Greyrate) are insignificant. This means that Vietnam's grey (pressure-sensitive) institutions are less likely to be concerned about firms' sustainability (relating to their size and age) to deliver the investment decision.
Regarding the control variables, higher Tobin's Q, operating performance (ROA), and market share (Mkshare) are positively related to private institutional investors, whereas having the state as the ultimate controlling shareholder is negatively related to them. However, Column (4) in Table 3 reveals that state-owned institutional investors prefer to invest in state-owned firms, unlike privately owned institutional investors. Moreover, state-owned institutions are likely to hold shares in higher leverage (LEV) firms, whereas privately owned institutions prefer lower leverage (LEV) firms. In terms of grey and independent ownership, both types of companies are negatively associated with a firm with higher leverage (LEV). Independent institutions are positively related to Tobin's Q and negatively related to LEV and STATE. On the other hand, ownership by grey institutions was positively associated with STATE and Mkshare.
The R-squared values ranging from 25.88% to 81.04% in this model are relatively better than those obtained from previous studies; for example, Guo and Platikanov (2019) reveal R-squared values of 32.2%.

Institutional ownership and firm performance
After identifying institutional investors' preferences for Vietnamese listed companies, this study examines the effect of different categories of institutional ownership on firm performance, including grey, independent, privately owned, and state-owned institutions. We test Hypotheses 3 and 4 by applying pooled OLS (single-equation test) and 3SLS models (endogeneity treatment).

Single equation tests
Results of using the pooled ordinary least squares method to examine the effect of institutional ownership on firm value are shown in Table 4. Tobin's Q is the dependent variable that measures firm performance. All estimations contain time-fixed effects (�δ t ).
In Group1, the results in Column (1) indicate that there is a significantly positive relationship between independent institutions (Indrate) and Tobin's Q, while grey institutions are negatively associated with Tobin's Q through the coefficients of 0.12355 and −0.96126, respectively, at the 10% statistical significance level. In Group2, Column (2) shows that both privately owned (Private) and state-owned institutional ownership (Starate) are positively correlated with Tobin's Q. These findings are highly consistent with Hypothesis 3 and a part of Hypothesis 4, which states that the effect of grey institutions on firm performance is negative. In contrast, the effect of independent, privately owned, and state-owned institutions on firm performance is positive. These results coincide with the conclusions of Cornett et al. (2007), Ferreira and Matos (2008), Lin et al. (2009), and Guo and Platikanov (2019). However, there is no evidence to declare that privately owned institutions have more effects on firm performance than state-owned institutions.
Management pressure is essential in institutional investors' impact on the market performance of listed Vietnamese companies. The controlling incentive of an institutional investor is derived from being independent, regardless of whether they are state-owned or privately owned. This study is the first to examine the relationship between the two dimensions of institutional ownership (grey versus independent and private versus state), corporate governance, and firm performance in Vietnam.
Concerning other explanatory variables, Table 4 shows that the number of boards of directors (Adnum), company size (SIZE), and return on assets (ROA) are positively related to Tobin's Q. In contrast, there is a negative relationship between the variables (STATE and CEO Duality) and Tobin's Q. Moreover, companies with larger institutional investors and monitoring each other relate to Tobin's Q positively.
All estimations reported in Model (2) are statistically significant. The R-squared values for Group1 and Group2 in Table 4 are 17.66 and 18.67%, respectively. Guo and Platikanov (2019) argue that institutional ownership and firm value can be simultaneously interdependent. This study uses Tobin's Q to measure market performance. Tobin's Q could impact investors' decisions on whether they should invest in the company stock and how large the investment should be. Moreover, Tobin's Q could also be a crucial factor for current investors' decisions on whether to change the size of their existing position. On the contrary, Tobin's Q might be affected by the company's ownership structure in general and by large institutional shareholders in particular. Therefore, major shareholders' institutional ownership and monitoring roles influence managerial decisions and firm performance.

Endogeneity and treatment
A three-equation system could be applied to address the potential simultaneity and endogeneity, where Tobin's Q, Group1 (grey and independent institutions), and Group2 (privately owned and stateowned institutions) are the dependent variables in each of the individual equations in the system.
Results from the 3SLS method, where the dependent variables are Tobin's Q and Group1 are presented in Table 5. All estimations contain time-fixed effects (�δ t and�θ t ).
The first system of multiple equations allows this study to attain separate estimates for the effect of grey and independent institutional ownership on Tobin's Q. Following Table 5, the estimates reported in Column (1) specify that independent institutional ownership is positively and significantly related to firm performance, with a coefficient of 0.434 and z-statistic of 5.18. In contrast, grey institutional ownership has a negative and insignificant effect on firm performance, with a coefficient of −0.159 and a z-statistic of −1.03. Table 6 presents the three-stage least-squares method results, where Tobin's Q and Group2 are the dependent variables. All estimations contain time-fixed effects (�δ t and�μ t ).
The second system of multiple equations is adopted to investigate the effect of privately owned and state-owned institutional ownership on Tobin's Q. Following Table 6, the estimates reported in Column (1) indicate that the coefficients and z-statistics of privately owned institutional ownership on Tobin's Q are 3.152 and 3.77, respectively, whereas those of state-owned institutional ownership are 1.946 and 3.47, respectively. This result suggests that both private and state institutions have significant and positive impacts on firm performance. The results in Table 5 and Table 6 conclude that pressure sensitivity to management, rather than the nature of institutional investors (privately owned or state-owned), is vital for institutional ownership's effect on listed companies' performance in Vietnam. By contrast, both private and state institutions have a positive impact on firm performance. This finding is equivalent to Guo and Platikanov's (2019) evidence of the positive effects of independent (pressure-insensitive) financial investors, comprising investment advisory and venture capital investors for Chinese companies.

Robustness test of institutional investors' preferences
The results in Table 3 indicate some corporate governance and business features preferred by institutional investors in Vietnam. Specifically, they are less likely to hold firms with larger board sizes and CEO duality and are ultimately controlled by the state. This section presents robustness checks of institutional investors' preferences for corporate governance and firm characteristics. We used principal component analysis (PCA) to measure the quality of corporate governance by computing the corporate governance score. Equation (6) provides a corporate governance system, implying that a higher index value indicates stronger corporate governance performance.
where BMEET represents the number of board meetings, EDR denotes the ratio of executive directors to board size, and WSIZE denotes the number of women on the board of directors. Table 7 reports robustness results from testing the association between determinants of institutional ownership and the corporate governance index measure in equation (7) below.
We also find evidence to prove Hypothesis 1, which states that institutional investors in Vietnam always favor firms with strong corporate governance (especially independent and state-owned institutional ownership). In addition, Table 7 reinforces the findings on institutional investors' preferences for some firm characteristics, including larger size, older age, and members of the VN30 index (Hypothesis 2).

Conclusion and recommendations
By analyzing the data on institutional ownership of Vietnamese listed companies, this study examines institutional investors' preferences for specific corporate governance features and company characteristics and investigates the impact of institutional ownership on firm performance. From the names and types of large investors in each listed company in Vietnam, financial institutions are manually identified to classify them into one of the following categories: (1) grey or pressure-sensitive institutions, and (2) independent or pressure-insensitive institutions. In particular, banks and insurance companies are classified as grey institutions based on their significant business relations with the firms they invest in. On the other hand, securities companies and venture capital firms are classified as independent institutions concerning the insignificantly related business with the firms in which they invest. Private and state ownership were also adopted in this research. Thus, institutional ownership is separated into four categories: (1) ownership by independent institutions and grey institutions and (2) ownership by state institutions and private institutions.
The findings in this study indicate that while institutional investors are less likely to hold shares of companies with larger board sizes, CEO duality, and ultimately control by the state (except for state-owned institutions' perspective), the effect of their ownership on firm performance depends on whether they are grey or independent institutions. Additionally, PCA analysis to compute corporate governance performance reinforces that institutional investors prefer firms with stronger corporate governance quality. In Vietnam, independent institutional investors monitor the company they invest in more effectively than grey institutional investors do. They can significantly influence management decisions and improve shareholder value economically. Further analysis indicates that the positive association between institutional ownership and Tobin's Q is mainly due to independent institutional ownership rather than any other type of institution. On the other hand, grey institutional ownership is either negatively or insignificantly related to firm performance concerning conflicts of interest, as they have a potential related business relationship with the invested companies. These results were modified to overcome the potential endogeneity of institutional ownership and firm performance, transformations of the dependent variable, and different estimation methods.
The findings of this study also suggest recommendations for listed companies and authorities. First, managers should identify the company size suitable for their company's current business context. This can be achieved by determining the economies of scale for their products and services and the business operation system of the entire company. Next, the board of directors should implement practical measures to enhance their companies' corporate governance features. The company's charter should declare that there must be a separation between the CEO and the chairman of the member's council. While having the CEO and the chairman distinct, the top manager can independently deliver strategic decision-making and policy implementation. This also prevents the act of expropriating wealth from large shareholders. Moreover, companies should increase their number of large institutional investors as a controlling mechanism. Shareholders might gain better protection in companies where the largest shareholders closely monitor each other. Besides one or two strategic investors, the board of directors can invite large shareholders with experience in many related business fields. They can apply their knowledge to the R&D process and play a monitoring role among existing shareholders to ensure the transparency of the board of directors.
Following the empirical results, there are several modifications that the government should impose to enhance the Vietnamese economic environment (the emerging market) to a more transparent and efficient one. First, because pressure sensitivity to management, rather than the nature of institutional investors, is essential for the effect of institutional ownership on listed companies' performance in Vietnam, dependent financial institutions can positively influence firm value. The Vietnamese government should promulgate regulations and policies that can actively encourage the development of independent financial institutions, such as pension funds, asset management, investment advisory, and venture capital firms. Additionally, domestic and foreign investments should be promoted to increase stock market efficiency. The government can achieve this by designing policies to reduce capital costs that improve the nation's economic wealth. Next, regulatory agencies (Ministry of Finance, State Securities Commission, Ha Noi, and Ho Chi Minh Stock Exchanges) should study and apply the standards and principles to improve corporate governance in developed countries in the context of the Vietnamese economy. Thus, supplementing and completing regulations on corporate governance and rules on disclosure of information for listed companies can enhance transparency in the financial market's business environment. For example, according to Article 152, Sub article 2, the Enterprise Law 2014, the chairman of the board of directors of any joint-stock company over 50% of voting shares held by the state may not concurrently hold the position of the general director. However, the government can significantly modify this regulation by separating the roles of the chairman and CEO in many different areas and circumstances. The issue of this regulation is one way to prevent the largest shareholders from expropriating the wealth of other stakeholders. Finally, being controlled by the state can hurt firm performance. If the state dominates a corporation, managers may gain more freedom and authority in decision-making. Thus, they can quickly seek internal benefits that result in inefficiencies in company operations. This study recommends that the Vietnamese government accelerate the privatization process in each sector investment and gradually reduce the state's presence as a controlling shareholder to attract domestic and foreign investment strategically. Enterprise equitization can be promoted by issuing decrees that eliminate difficulties and provide specific roadmaps and solutions to conduct restructuring activities efficiently.